A snapshot of the AI bubble market by UNA Coventry volunteer Adreen Fernando who is a student of Criminology and Sociology at Coventry University.

Not only does Generative Artificial intelligence (IBM, 2023) endanger our beautiful planet, but it also alters human existence. On one hand, billionaires and experts (Bhati, 2025; The Economic Times, 2023) issue warnings and risk assessments that AI might be used to stimulate conflict and wars, which are short-term consequences! Short-term. AI can enable misinformation, deep fakes, and cyber disruption that destabilise states. Militaries (China, Israel, Russia, etc) are already employing AI for surveillance, target identification and semi-autonomous strike systems (Bruun & Palayer, 2025).

So what’s another short-term consequence of the current AI boom? The stock market.

The staggering rise of artificial intelligence has sparked one of the most significant investment booms in modern history, with tech giants and startups alike racing to dominate what many consider to be the next technological revolution. As Sundar Pichai, CEO of Alphabet (Google’s parent company), told the BBC in a recent interview, while AI represents an “extraordinary moment” of technological advancement, the current investment climate contains clear “elements of irrationality” (Clun & Islam, 2025). The genuine transformation and speculative excess defines the current debate about whether we are in an AI Bubble, and if so, how might it compare to historical crashes like the 2008 financial crisis or the dot-com bubble of 2000.

The term “AI bubble” refers to the theorised stock market bubble growing amid the current AI boom, characterised by concerns that leading tech firms are involved in circular investments that artificially inflate their stock values (Wikipedia, n.d.). With Nvidia becoming the first company to reach a $5 trillion market valuation in 2025 and AI-related enterprises accounting for roughly 80% of gains in the American stock market over the past year, questions about sustainability and speculation have reached a fever pitch (Nishant & Singh, 2025).

This report contends that Open-Source Intelligence (OSINT) provides a critical, evidence-based framework to move beyond speculation and analyse the underlying geopolitical realities of the so-called “AI bubble.” By systematically collecting and analysing publicly available data, we can assess national strategies, supply chain vulnerabilities, and market signals to determine whether current valuations are built on sustainable foundations or geopolitical speculation. Our analysis identifies key pressure points, including the US-China tech competition, semiconductor supply chain concentration, and the alignment of massive capital expenditure with genuine strategic demand, which will be decisive factors in the market’s future trajectory.

THE NEED FOR A GEOPOLITICAL, HOLISTIC LENS

The debate around an “AI bubble” often centres on financial metrics alone: price-to-earnings ratios, market capitalisation, and venture-capital flows. However, AI isn’t merely a market sector; it’s a foundational technology that many nations regard as central to future economic and military dominance. This geopolitical dimension is a primary driver of investment and systemic risk. OSINT, understood as the discipline of collecting and analysing publicly available information for intelligence and strategic analysis, is uniquely suited to disentangle this complex web. A structured OSINT methodology can reveal, beyond speculative hype, the underlying infrastructure, power dynamics, and systemic vulnerabilities shaping the AI boom, which can help identify early warning indicators of market stress.

OSINT FRAMEWORK FOR GEOPOLITICAL ANALYSIS

A multi-layered OSINT approach enables a holistic view of the AI landscape by triangulating data from government/policy, corporate/market, and technical/supply-chain domains.

Government & Policy Intelligence

1. National AI Strategies: Tracking official strategy documents, white papers, and state budget allocations across major powers (e.g., U.S., China, EU) can reveal real long-term commitment to AI infrastructure and capabilities.
2. Regulatory Tracking: Monitoring the implementation of AI-relevant regulation (such as export-control regimes, trade restrictions, or AI governance initiatives) offers insight into which technologies and supply-chain nodes may face strategic constraints.
3. Legislative Monitoring: Parliamentary debates, congressional hearings, and committee reports can offer early signals of regulatory shifts, national security concerns, or new funding allocations that might influence AI-market trajectories.

Corporate & Market Intelligence

• Financial Disclosures: Analysing public filings (e.g., SEC 10-K/10-Q in the U.S., equivalent filings elsewhere) enables tracing of capital expenditure, R&D spending, and revenue attribution, which helps distinguish hype from realised income.
• Earnings Call Analysis: Employing natural language processing or content analysis on earnings-call transcripts can detect management sentiment, timelines, and contract/backlog announcements.
• Patent and Research Analysis: Using databases such as USPTO, WIPO, academic preprint servers (e.g., arXiv), and public research registries can map flows of innovation, identify key players, and signal which research is being commercialised or weaponised.

Supply Chain & Infrastructure Intelligence

• Trade Analysis: Using international trade databases (e.g., UN Comtrade) and national customs statistics to map flows of critical components: semiconductors, lithography tools, rare-earth elements, and other critical materials.
• Infrastructure Mapping: Using satellite imagery, public permit filings, energy consumption statistics, and data-centre registry information to track the build-out of AI infrastructure globally (data centres, fab-plants, packaging facilities).
• Talent Flow Analysis: Monitoring public information on researchers (academic publications, LinkedIn / professional-network profiles, conference participation) to detect migration patterns, hiring surges, or talent clustering across countries and firms.

GEOPOLITICAL APPLICATIONS AND FINDINGS

The US-China Tech Decoupling

• Findings: Public documents show tightening export controls on advanced chips and chip-making equipment destined for China. For example, since 2022 the U.S. government has explicitly targeted frontier AI-relevant microprocessors, memory chips, lithography systems, and electronic-design tools for export restrictions, seeking to throttle Chinese access to advanced manufacturing capabilities (Harithas & Schumacher, 2024).
  • Simultaneously, publicly available policy documents and corporate filings show that Chinese authorities and firms are intensifying efforts to achieve semiconductor self-sufficiency (e.g., boosting domestic foundry capacity, subsidising chip manufacturing, and investing in domestic R&D). Though publicly available documentation is more fragmented, evidence from multiple trade analyses and industry signals points to this state-backed push.
• Bubble Implication: This decoupling creates a duplicated, inefficient ecosystem. It also shrinks the addressable global market for many Western AI firms (if China becomes a largely self-sufficient tech bloc), thereby undermining growth assumptions baked into valuations predicated on a single, integrated global AI market.

  THE SEMICONDUCTOR SUPPLY CHAIN VULNERABILITY

The concentration of advanced semiconductor manufacturing poses a systemic risk.

• Findings: Recent market data shows that one company, Taiwan Semiconductor Manufacturing Company (TSMC), dominates the global pure-play foundry market. In Q2 2025, TSMC captured approximately 71% of total pure-play foundry market share, according to industry analysts, with much of the surge driven by demand for AI and high-performance-computing chips utilising its advanced 3nm, 4nm and 5nm process technologies (SemiMedia, 2025).
  • Additionally, the broader “Foundry 2.0” market, which includes not only pure-play foundry but also advanced packaging, photomask manufacturing, OSAT, and other support services, has grown rapidly (revenues rising by 13% YoY in Q1 2025), driven by AI/HPC demand (Counterpoint, 2025).
  • Separately, ASML, the Dutch firm supplying extreme-ultraviolet (EUV) lithography systems, essential to making the most advanced chips, remains effectively a monopoly in that critical niche (Schneider, 2025).
• Bubble Implication: Such concentration, both at the foundry level (TSMC) and at the equipment-supplier level (ASML), creates systemic risk: any geopolitical disruption (e.g., conflict, export-control escalation, supply-chain blockade) centred on key nodes (e.g., Taiwan, the Netherlands) could instantly halt the AI infrastructure build-out. That underlying fragility may not be fully priced into financial valuations, implying a significant tail risk.

STRATEGIC INVESTMENT VS. MARKET DEMAND

OSINT helps distinguish between state-driven investment and organic market demand.

• Findings: The surge in foundry and related manufacturing revenues in 2024–2025 correlates strongly with reported increased demand for AI and HPC chips, suggesting that AI infrastructure, not only consumer-electronics demand, is a major driver of growth (Counterpoint, 2025). Meanwhile, public regulatory documents and export-control regimes confirm that governments are actively using trade and technology policy as a lever: restricting access to cutting-edge chips or chip-making tools for certain jurisdictions (Harithas & Schumacher, 2024).

• • However, there is also evidence (via recent academic research) that hardware-centric controls may be imperfect. Firms may adapt by running AI workloads on less advanced chips, optimising software or model efficiency, or using alternative architectures. For example, recent work demonstrates how less-capable hardware combined with algorithmic efficiency can partially circumvent hardware constraints imposed by export-controls (Gupta et al., 2024)

• Bubble Implication: This suggests that part of the AI-investment cycle is driven by strategic/geopolitical priorities (state funding, national tech-leadership), not just by organic enterprise demand. If commercial adoption or profitability fails to catch up with hype, or if state-driven demand slows, there may be a sharp correction. The risk is that valuations might reflect overoptimistic assumptions about widespread commercial demand, rather than grounded expectations of realised revenues.

EARLY WARNING INDICATORS OF A GEOPOLITICAL BURST

Based on the above OSINT-grounded framework, the following indicators deserve close monitoring:

• Escalation of Export Controls: Tightening of export restrictions under instruments like the U.S. CHIPS Act, or reciprocal controls from other powers (e.g., EU, Netherlands), especially targeting lithography tools, high-end chips, or packaging technology.
• Divergence in Technical Standards / Supply Chains: Emergence of incompatible AI stacks, fabrication standards, or supply-chain ecosystems between major blocs (e.g., U.S + allies vs. China + others), fragmenting the global market.
• Sharp Reduction in Cross-Border Research / Collaboration: A measurable decline in co-authored AI papers, cross-national research projects, or international talent exchange — indicating siloing of innovation.
• Slowdown in CapEx / Debt-Financed Infrastructure Investment: Public filings (e.g., securities filings) showing major tech firms delaying or cancelling data-centre or chip-fabrication build-outs due to rising debt costs or insufficient demand.
• Geopolitical Flashpoints: Heightened political or military tension in critical regions (e.g., the Taiwan Strait), which could threaten core manufacturing nodes and supply-chain continuity.

ADDRESSING A MISCONCEPTION

It’s excellent and sharp to question whether this AI Bubble phenomenon is a Ponzi Scheme. While they share similarities, the fear that later investors are funding earlier ones without underlying value creation are different in their structure, intent, and mechanism. The AI Bubble lacks criminal intent and is based on a genuine technological breakthrough.

However, concerns over this are 100% correct. The market is currently valuing these companies based on a future revenue stream that doesn’t yet exist at scale. The ‘circular’ flow of capital within the tech ecosystem is a major red flag that the hype has outpaced the tangible economic value creation. If the massive external customer demand doesn’t materialise to justify the billions spent on infrastructure, the bubble will deflate. This would be a classic market correction based on failed expectations, not the criminal collapse of a fraudulent scheme. The “greater fools” will run out, and the companies built on hype rather than solid business models will disappear.

CONCLUSION

The question of an AI bubble cannot be answered by financial analysis alone. The market is being fundamentally shaped by geopolitical forces. An OSINT-led approach demystifies this relationship, replacing speculation with data. The current evidence points to a market fueled by both genuine technological transformation and significant geopolitical speculation. While the long-term potential of AI is undeniable, the path is fraught with risks rooted in supply chain concentration, tech nationalism, and the potential misallocation of capital based on strategic rather than economic imperatives. The AI bubble mirrors 2008 in terms of hype, speculative investment, and FOMO-driven market psychology. The difference is that AI has tangible technological value and strategic significance, meaning even a correction may not erase the progress made. Still, systemic vulnerabilities (supply chains, geopolitical friction, over-leveraged companies) mean the AI bubble could lead to sharp market adjustments if expectations diverge from reality. Continuous OSINT monitoring of the indicators outlined in this report is essential for any stakeholder seeking to understand the true stability, or fragility, of the AI-driven market.

REFERENCES

AI Bubble. (n.d.). AI Bubble. https://en.wikipedia.org/wiki/AI_bubble

Baburajan, R. (2025). Big tech to spend trillions on AI infrastructure by 2030: McKinsey, Gartner, Bain & Citigroup forecast massive growth. Infotechlead. https://infotechlead.com/artificial-intelligence/big-tech-to-spend-trillions-on-ai-infrastructure-by-2030-mckinsey-gartner-bain-citigroup-forecast-massive-growth-91368?utm_source=chatgpt.com

Bhati, D. (2025). AI godfather warns killer robots could make war easier to start, not safer to fight. India Today. https://www.indiatoday.in/technology/news/story/ai-godfather-warns-killer-robots-could-make-war-easier-to-start-not-safer-to-fight-2778692-2025-08-29?utm_source=chatgpt.com

Brunn, L., & Player, J. (2025). Artificial intelligence and international peace and security. Stockholm International Peace Research Institute.

Counterpoint Research. (2025, June 24). Global semiconductor foundry 2.0 market’s Q1 2025 revenue jumps 13% YoY driven by AI chip demand. Counterpoint Research. https://www.counterpointresearch.com/insight/post-insight-global-semiconductor-foundry-20-markets-q1-2025-revenue-jumps-12-yoy-driven-by-ai-chip-demand/?utm_source=chatgpt.com

Economy.ac. (2025, May). TSMC and ASML, the “serial chip market conquerors” and “national security players”. Economy.ac. https://economy.ac/news/2025/05/20250549697?utm_source=chatgpt.com

Gupta, R., Walker, L., & Reddie, A. W. (2024). Whack-a-Chip: The futility of hardware-centric export controls. arXiv. https://arxiv.org/abs/2411.14425?utm_source=chatgpt.com

Harithas, B., & Schumacher, A. (2024). Where the chips fall: U.S. export controls under the Biden administration from 2022 to 2024. CSIS. https://www.csis.org/analysis/where-chips-fall-us-export-controls-under-biden-administration-2022-2024?utm_source=chatgpt.com

IBM Research. (2023). What is generative AI. IBM Research. https://research.ibm.com/blog/what-is-generative-AI

Islam, F., & Clun, R. (2025). Google boss says trillion-dollar AI investment boom has ‘elements of irrationality’. BBC News. https://www.bbc.com/news/technology-67312345

Nishant, N., & Singh, R. (2025, October 29). Nvidia hits $5 trillion valuation as AI boom powers meteoric rise. Reuters. https://www.reuters.com/business/nvidia-poised-record-5-trillion-market-valuation-2025-10-29/?utm_source=chatgpt.com

SemiMedia. (2025). TSMC strengthens foundry dominance with 71% market share in Q2 driven by AI demand. SemiMedia. https://www.semimedia.cc/20022.html?utm_source=chatgpt.com

Soni, A., & Sophia, D. M. (2025, October 29). Microsoft’s massive AI spending draws investor concerns as cloud business booms. Yahoo Finance. https://finance.yahoo.com/news/microsofts-massive-ai-spending-draws-224510514.html?utm_source=chatgpt.com&guccounter=1&guce_referrer=aHR0cHM6Ly9jaGF0Z3B0LmNvbS8&guce_referrer_sig=AQAAAA1dZ3mPantne-qKcbiH81iR4kROnCuUe_CjnpICod7Ux6JZuklnUo2Jj8hlTOFUtmn3mhucv8JpvdD6Bm8fWTE5mSVPLSZNDICAfYRehjWKTVdrEG3p2phgxS_Qk4CggnX0s8Z6jr5Yb90tzVW1JtYVcHNPiy3MI-2mvANbL7lM

The Economic Times. (2023). AI may begin war, increase nuclear arsenal, warns ex-Google engineer Blake Lemoine. https://economictimes.indiatimes.com//news/international/us/ai-may-begin-war-increase-nuclear-arsenal-warns-ex-google-engineer-blake-lemoine/articleshow/106300767.cms?utm_source=contentofinterest&utm_medium=text&utm_campaign=cppst

Wright, I. (2025). ChatGPT energy consumption visualised. Business Energy UK. https://www.businessenergyuk.com/knowledge-hub/chatgpt-energy-consumption-visualized/#:~:text=With%20this%20information%2C%20we%20established,consumption%20is%2039.98%20million%20kWh